The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs shell operations for skill installation and management, including creating symbolic links (ln -s), managing version control (git commit, git revert, git push), and executing a local diagnostic script (bash scripts/redlight-scan.sh). These operations are confined to the user's home directory and skill-specific paths.
[SAFE]: The skill implements robust security and quality guardrails, including mandatory human-in-the-loop checkpoints (🔴 CHECKPOINT) before data is pushed to remote repositories. It explicitly prohibits dangerous commands such as rm -rf and git reset --hard through a dedicated blacklist and includes a 'red light scan' to identify platform-specific hardcoding.
[EXTERNAL_DOWNLOADS]: The skill references best practices from various external GitHub repositories, including those from Anthropics, and suggests using standard ecosystem tools like npx for managing skill updates. These references are used for architectural guidance and tool management.

laohan-skillcreator